Restricting access for CDP services that create their own security groups on AWS
The security groups that you select to use during environment registration are only used for the Data Lake, FreeIPA, Data Hubs, and Operational Databases running in that environment. The Kubernetes-based CDP services (Data Engineering, Data Flow, Data Warehouse, and Machine Learning) create their own security groups with rules that should be restricted separately.
The following table explains where and when you can restrict these rules:
|CDP service||Type of access that can be restricted||When and where to restrict||Link to related documentation|
|DataFlow||Admin access to the Kubenetes API Server endpoint can be restricted.
End user access can be restricted.
|Restrict admin access to Kubernetes endpoints during or after enabling DataFlow
via the Kubernetes API Server Endpoint Access setting.
Restrict end user access to the the DataFlow endpoints during or after enabling DataFlow via the Load Balancer Endpoint Access setting.
|Enabling DataFlow for an environment|
|Data Engineering||Admin access to Kubenetes endpoints can be restricted.
End user access can only be restricted manually from the AWS management console.
|Restrict admin access to Kubernetes endpoints during enabling Data
Engineering via the Whitelist IPs parameter.
Restrict end user access manually from the AWS management console.
|Enabling Cloudera Data
|Data Warehouse||Both admin access to Kubernetes endpoints and end user access are always set to the
same range that can be set in environment activation settings.
While the access to the Kubernetes endpoints is a combination of the Cloudera Control Plane’s CIDR and your CIDR provided in environment activation settings, the access to the end user access points (JDBC, UI) is only your CIDR provided in environment activation settings.
|In Data Warehouse environment’s activation settings.||Restricting access to endpoints in AWS
|Machine Learning||There are two separate options, one for admin access to Kubernetes endpoints and another for end user access.||During ML workspace provisioning, under Network Settings:
||Provisioning ML Workspaces|